A crucial element in the development of effective prophylactic strategies for AIDS is an experimental animal model in which the course of immunodeficiency virus infection parallels the pathogenesis of the human disease. SIV infection of macaques is a relevant model. It induces an immunodeficiency syndrome in infected macaques that is remarkably similar to human AIDS. Therefore, candidate vaccines can be evaluated not only for their ability to prevent infection but also for their ability to prevent AIDS. The major vaccine effort within the laboratory has been an evaluation of the highly attenuated vaccinia virus Ankara (MVA) strain as a recombinant vector. Three recombinant MVA viruses were generated that expressed either gag-pol alone, env alone, or a combination of gag-pol and env. A cohort of rhesus macaques were immunized to evaluate immunogenicity and protective efficacy after intravenous challenge. All macaques became infected following intravenous challenge with SIVsmE660. However, plasma viremia in each of the groups immunized with MVA-SIV recombinants was significantly reduced as compared to the group vaccinated with nonrecombinant MVA and survival was prolonged. These data demonstrate that vaccination with MVA-SIV recombinants results in significant protection from high viremia and AIDS. However, longterm followup demonstrated that the vast majority of these immunized macaques have progressed to AIDS. Evaluation of gag-specific CTL was performed in collaboration with Dr. Norman Letvin, utilizing macaques expressing the MamuA*01 MHC Class I haplotype. Four MamuA*01+ macaques were immunized with the MVA-gag-pol recombinant virus and an additional two macaques were immunized with MVA. Macaques immunized with MVA expressing gag-pol developed CD8+ CTL. Challenge of these macaques with SIVsmE660 resulted in a rapid and substantial anamnestic CTL response. Viral load set point was reduced in the macaques immunized with MVA-Gag-pol as compared to controls, however, the difference was not statistically significant. We evaluated the MVA strategy using a different challenge model, SHIV/89.6P infection of rhesus macaques. Four MamuA*01+ macaques were immmunized with MVA expressing SIVmac239 gag-pol and MVA expressing HIV/89.6 env. Following challenge with pathogenic SHIV/89.6P, a significant reduction in set point viremia and partial protection from CD4 lymphocyte depletion was observed when compared with animals immunized with MVA nonrecombinant. This degree of protection is only slightly less robust than that observed in macaques immunized with cytokine augmented DNA and DNA prime-MVA boosted macaques. This contrasts with the far less robust protection observed in macaques immunized with a comparable regimen following SIV challenge. During primary viremia in SHIV-infected macaques, massive elimination of CXCR4 anive CD4+ T cells occurred. In contrast, CCR5+ memory CD4+ T cells were selectivley depleted in rapidly progressing SIV-infected macaques. Thus SHIV and SIV target different subsets of CD4+ T cells. These differences explain the different pathogenesis of SIV and SHIV. Importantly, in the context of developing an effective vaccine, regimens that suppress SHIV might not protect monkeys against SIV or humans against HIV. In a collaborative study with Nancy Haigwood (SBRI), the role of passively transferred antibodies were studied in the SIV/macaque model. Previous studies showed that postinfection passive treatment with polyclonal immune globulin with high neutralizing titers against SIVsmE660 (SIVIG) significantly improved the 67-week health of SIVsmE660-infected rhesus macaques. SIVIG treatment delayed the de novo production of envelope (Env)-specific antibodies by 8 weeks. Longterm follow-up of these animals showed that differences in disease progression were significant at 5 years postinfection. Macaques that maintained less than 1000 virus particles per ml of plasma had delayed disease onset. All macaques that survived beyond 18 months had measurable Gag-specific cytotoxic T cells, regardless of treatment. Humoral immunity in survivors beyond 20 weeks was strikingly different in the SIVIG and control groups. Despite a delay in Env-specific binding antibodies, de novo production of neutralizing antibodies was significantly accelerated in SIVIG-treated macaques. Titers of de novo neutralizing antibodies at week 12 were comparable to levels achieved in controls by week 32 or later. Acceleration of de novo simian immunodeficiency virus immunity in the presence of passively transferred neutralizing antibodies is a novel finding with implications for postexposure prophylaxis and vaccines. PHYLOGENY OF SIV/HIV: The human immunodeficiency viruses, HIV-1 and HIV-2, are members of an extensive family of primate lentiviruses that appear to have their origins in African primates. Each of the human viruses presumably arose following cross-species transmission from a naturally-infected primate to humans. In the case of HIV-2, the precursor appears to be SIVsm from sooty mangabey monkeys (Cercocebus atys) whereas the origins of HIV-1 arose from SIVcpz from chimpanzees. The goal of this portion of the project is to molecularly characterize novel SIV isolates from wild-caught African monkeys more extensively. We initiated these studies by characterizing SIV isolates from three of the species of African green monkeys (vervets, grivets and tantalus), then SIV from Sykes monkeys and most recently SIV from lhoest monkeys (C. lhoesti). Studies in the Peeters' lab revealed that mandrills are infected with two distinct types of SIVs, one being the previously characterized virus now named SIVmnd-1 which is related to virus of the SIVlhoest species. The other appears to be a recombinant between SIVmnd-1 and SIV from drills (SIVdrl). Both SIVmnd-2 and SIVdrl have regions of homology with SIV from redcapped mangabeys (SIVrcm) but neither SIVrcm nor SIVdrl have been fully characterized. Recently we cloned and sequenced the entire genome of SIVrcm isolated from a wild caught redcapped mangabey from Nigeria and demonstrated that this virus is a complex recombinant. Comparison of SIVmnd-2 with SIVrcm revealed that SIVmnd-2 was a mosaic genome, highly related to SIVrcm in the 5' portion of the genome and to SIVmnd-1 in the 3' portion of the genome. Thus it appears that SIVrcm while in itself a recombinant is also the source of a portion of the SIVmnd-2 genome. These data are suggestive of cross-species transmission and recombination in the evolution of these viruses. We also characterized a SIVdrl and novel SIVmnd-type 2 isolate. These studies revealed that both of these viruses are recombinants between a SIVrcm-like virus and a SIVlhoest like virus and share a common recombination point. These data are consistent with a complex history of recombination and cross-species transmission involving SIV in drills, mandrills, redcapped mangabeys and L'hoest monkeys.